Simplified Methods of Evaluating the Redundancy of Twin Trapezoidal Box Girder Bridges

According to the AASHTO Load and Resistance Factor Design (LRFD) Bridge Design Specifications, a bridge is defined to be fracture critical when a failure of a tension component will result in the collapse of the bridge. In the case of a twin box girder bridge, the tension flanges in the positive moment portion of the bridge, as well as the webs, are considered to be fracture critical elements. Due to this classification, those bridges are subjected to stringent inspections at least every two years. Those inspections are crucial for ensuring the safety of the bridge, yet are expensive and time consuming. Multiple cases of FCBs (Fracture Critical Bridge) that have experienced a fracture in one of their elements without collapse have encouraged owners of those bridges to question the validity of AASHTO’s requirements. The Texas Department of Transportation is interested in indentifying when a fracture of an element could lead to a catastrophic collapse of a bridge. A better understanding of fracture critical bridge behavior may allow TxDOT and other state DOTs to reduce the frequency of the inspections, which could potentially reduce the cost of an otherwise attractive bridge design.

The goal of this research project is to determine the level of redundancy of twin box girder bridges. Simplified analytical methods and guidelines that will conservatively estimate the behavior of such bridges will be presented in this thesis. Those guidelines will be one of the tools that an engineer in practice could use to determine if a bridge is prone to collapse following the failure of a fracture critical component. A full-size bridge has been constructed at the Ferguson Structural Engineering Laboratory to test the response of these systems following a simulated fracture. A series of tests were conducted to determine the response of the bridge in the event of a tension flange fracture. The results provided important information for the development of the simplified methods.

The FSEL test bridge performed extremely well throughout all the testing and supported a load of over four times the AASHTO design truck load. Several elements contributed to create alternative load paths that could sustain the entire applied load with a full-depth fracture of one of its two girders. The large section of the concrete railing above the fractured girder acted as an inverted beam and transmitted a portion of the load back to the supports once the expansion joint closed due to the downward deflection of the bridge. The concrete deck acted as a shear diaphragm and also transferred significant loads in both horizontal directions. Because the performance of the test bridge far exceeded the AASHTO criteria, and because this behavior can be computed using the simplified methods presented in this thesis as well as through detailed finite element models, consideration should be given to revising the current AASHTO specifications and to developing alternate inspection and maintenance requirements that accurately reflect the redundancy available in various types of fracture critical bridges.